The Association Between Actigraphy-Derived Behavioral Clusters and Self-Reported Fatigue in Persons With Multiple Sclerosis: Cross-sectional Study

Abstract

BackgroundPersons with multiple sclerosis frequently report increased levels of fatigue and fatigability. However, behavioral surrogates that are strongly associated with self-reports are lacking, which limits research and treatment. ObjectiveThe aim of this study was to derive distinct behavioral syndromes that are reflected by self-reports concerning fatigue and fatigability. MethodsWe collected actigraphic data of 30 persons with multiple sclerosis over a period of 1 week during an inpatient stay at a neurorehabilitation facility. Further, participants completed the German fatigue severity scale. A principal component analysis of actigraphic parameters was performed to extract the latent component levels of behaviors that reflect fatigue (quantity of activity) and fatigability (fragmentation of activity). The resulting components were used in a cluster analysis. ResultsAnalyses suggested 3 clusters, one with high activity (d=0.65-1.57) and low clinical disability levels (d=0.91-1.39), one with high levels of sedentary behavior (d=1.06-1.58), and one with strong activity fragmentation (d=1.39-1.94). The cluster with high levels of sedentary behavior further revealed strong differences from the other clusters concerning participants’ reported levels of fatigue (d=0.99-1.28). ConclusionsCluster analysis data proved to be feasible to meaningfully differentiate between different behavioral syndromes. Self-reports reflected the different behavioral syndromes strongly. Testing of additional domains (eg, volition or processing speed) and assessments during everyday life seem warranted to better understand the origins of reported fatigue symptomatology.